An Adaptive Model Following Control of Industrial Manipulators

Abstract

Using the adaptive model following control (AMFC) technique, a control system is developed to obtain high performance of an industrial manipulator that has wide variations in its payload and spatial configuration. The main advantage of this technique over other methods is simplicity in computation, which provides the capability of real-time control calculation with today's microprocessors. Moreover, it guarantees automatically the stability of the overall system without requiring any additional stability analysis. In order to demonstrate the capability of this technique, we have conducted a computer simulation for the Unimation PUMA 600 series manipulator. The simulation results agree well with the expected high performance of the control system: (1) computational requirements are much less than other existing ones, and (2) it provides the capability of adapting to large variations in payload and spatial configuration. Also presented is a comparison of the amount of its required computation with that of other methods.